Computabilità e traduzione interlinguistica: considerazioni sull’impiego dell’Intelligenza Artificiale in ambito traduttivo

This article contributes to heated debate around the multidisciplinary field of Artificial Intelligence by adopting a specific perspective: the translatological point of view. Machine Translation was the first application of computer technology in the humanities, both because of the general interest in the study of human language, given its strong identity connotation, and because of its high formalization. Through an overview of the architectures that have followed, the paper focuses on the fruitful dialogue between practice and theory, which triggered modern reflections on the mechanisms of translation as a decision-making and negotiation process. These include: the limits of the interlingual transfer, the entropy of information, the anisomorphism of natural languages, the mutual influence of language and cognition, the relevance of extratextual information and speakers’ cognitive horizons, the concept of textual fidelity. The deepening of these aspects opens up wider reflections, such as the thorny question of the relationship between natural and artificial intelligence, the essence of understanding, and the possible heuristic impediment constituted by imitation.